1) Add initialization for trx->xid in InnoDB trx pool
2) Add server UUID in the XID of InnoDB SR transactions
3) Modify trx_recover_for_mysql_callback() so that recovered InnoDB SR
transactions are not made visible to MySQL as recovered XA transactions.
4) Modify how SR transactions are detected and the wsrep_is_sr flag is
set for InnoDB SR transactions.
The method takes already locked lock object as an argument.
The caller must ensure that the lock object owns the underlying mutex.
Replaced homegrown wsrep::unique_lock with type alias from
std::unique_lock.
This patch adds the possibility to have client commands that do not
return results from DBMS. While processing such commands we must be
able to preserve errors until the next interaction with client.
Specifically if the transaction is bf aborted while processing such
a non-returning command, then we have to keep the deadlock error until
the client issues a command that may return the error.
To handle such cases, client_state::before_command() now takes
parameter keep_command_error. The DBMS is supposed set
keep_command_error true to instruct wsrep-lib to preserve errors (if
any) until the next command which sets keep_command_error false.
Dealing with a case where current client command does not return result.
Work in progress.
Fix typo and add assertions in keep_command_error()
Make keep_command_error a parameter to before_commit()
Fix comment about keep_command_error
Handle keep_command_error with s_must_abort in wsrep_before_command()
Fix unit test
* Added unit tests for transaction::xa_detach() and
transaction::xa_replay()
* Added unit tests for wsrep::xid
* Fixed minor issues pointed out by reviewer
This patch implments replaying for prepared XA transactions.
Replay may happen in the following cases:
1) The transaction is BF aborted in prepared state and is idle. In
that case, the transaction is handed over to rollbacker for replay.
2) The transaction is BF aborted while executing the
commit (i.e. before or after successful certification). In
which case the transaction replays itself from fragment storage.
3) The transaction is BF aborted while certifying its commit
fragment. This case is handled like replay for streaming transactions,
where the provider is directly involved and re-delivers the last
fragment.
Add support for detaching XA transactions. This is useful for handling
the case where the DBMS client has a transaction in prepared state and
disconnects. Before disconnect, the DBMS calls the newly introduced
client_state::xa_detach(), to cleanup the local transaction and
convert it to a high priority transaction. The DBMS may later attempt
to terminate the transaction through client_state::commit_by_xid() or
client_state::rollback_by_xid().
Also in this patch:
- Fix client_state::close() so that it does not rollback transactions
in prepared state
- Changed class wsrep::xid representation to hold enough information
so that DBMS can convert to its native representation
- Fix potential infinite loop in
server_state::find_streaming_applier(wsrep:xid&)
- Append SR keys on prepare fragment and make it pa_unsafe
- Handle one phase commit (simply fall back to two phase)
- Do not rollback prepared streaming clients in
server_state::close_orphaned_transactions()
which complicates diagnostics and debugging.
Don't ignore provider return codes and more verbose error logging for
sst_sent(), sst_received(), set_encryption_key() methods
Refs codership/wsrep-lib#127
After a local certification failure, commit order is released without
the setting the current position in DBMS. Which results in diverging
positions between provider and DBMS, if clean shutdown happens right
after local certification failure.
This patch add method set_position() to server_service class. So that
wsrep-lib can instruct DBMS to set the current position after local
certification failure releases commit order.
If the transaction fails during replay because of certification
failure, the provider will return control to applier without
terminating the transaction and transaction remains in
s_replaying.
Fixed transaction::after_statement() to handle the state changes
correctly if certification failure is returned from replay.
Replaying was extracted to separate private method from
after_statement(). Removed transaction::after_replay() as it
seems now unnecessary and it bypassed state change sanity checks.
Allowed replaying -> committed transaction transition to handle
the situation where DBMS allocates a new context and client_state
to do the replay.
when losing error voting:
- if NBO has failed locally (DBMS side), don't override original DBMS
error so it gets reported to the client
- otherwise, report "query interrupted" instead of "error during commit"
- Retry enter_toi() in poll_enter_toi() also for error_connection_failed
which means that the connectivity to the cluster has been lost,
a.k.a non-prim.
Certification keys are needed for NBO end to resolve dependencies
for the write sets which follow NBO end. Without keys the following
write sets do not detect dependency to NBO event and may start applying
too early.
If timeout option is give, enter_toi_local() and begin_nbo_phase_one()
retry provider::enter_toi() as long as return status indicates
certification failure, given timeout expires or the client is interrupted.
- High priority interface method to apply NBO begin, separate from
apply_toi() in order to avoid implementation to force interpreting
ws_meta flags.
- Method to put client_state into NBO mode when applying NBO begin.
The client_state will process in m_local mode.
- Unit tests for applying NBO
- Implemented calls to enter and leave NBO phase one and two
- Extended client_state mode checking to include m_nbo
- Changed client_state state and mode change sanity checks to
print a warning and assert() instead of throwing exceptions
to be more graceful in release builds.
Access to empty vector by using operator[] may cause stdlib++
assertions to fail. Replaced the vector data access to use data()
method which is valid operation even if the vector is empty.
Added unit test to reproduce assertion with empty mutable_buffer access.
Added -D_GLIBCXX_ASSERTIONS preprocessor option to debug builds
to catch standard library misuse.
Added gcc 8 and gcc9 into travis build matrix.
- Do not use [[noreturn]] with GCC older than 4.8.
- Use if/else instead of ternary operator in transaction
commit_or_rollback_by_xid() to avoid debug build failure with
GCC 4.7.
Remove methods `is_xa()`, `is_xa_prepare()`, and `xid()` from
client_service interface. Instead, transactions are explicitly
assigned their xid, through at start of XA.
* Add method `restore_prepared_transaction` to `client_state` class
which restores a transaction state from storage given its xid.
* Add method `commit_or_rollback_by_xid` to terminate prepared XA
transactions by xid.
* Make sure that transactions in prepared state are not rolled back
when their master fails/partitions away.
Class streaming_context can now keep track of database specific log
position for streaming. Previously, it was assumed that the log
position was based of off the buffer size which was filled in
client_service::prepare_fragment_for_replication(). However, for XA
the buffer may contain data that is not in the log. Therefore,
subsequent calls to prepare_fragment_for_replication() could not find
the appropriate log_position based on the certified_bytes(), which
would returns the total number of bytes return in the buffers that
were certified.
next_fragment is called outside the scope of a high_priority_switch,
and we may be in a different thread context then the streaming applier
was created in
Force fragment replication when XA transaction is prepared, with
prepare fragment. Commit fragment happens in before_commit().
Adjusted fragment removal, which cannot happen in atomically with the
executing transaction.
